Method fallbackExtraction

apps/sim/lib/file-parsers/pptx-parser.ts:78–108

(buffer: Buffer)

  }

  private fallbackExtraction(buffer: Buffer): FileParseResult {
    logger.info('Using fallback text extraction for PowerPoint file')

    const text = buffer.toString('utf8', 0, Math.min(buffer.length, 200000))

    const readableText = text
      .match(/[\x20-\x7E\s]{4,}/g)
      ?.filter(
        (chunk) =>
          chunk.trim().length > 10 &&
          /[a-zA-Z]/.test(chunk) &&
          !/^[\x00-\x1F]*$/.test(chunk) &&
          !/^[^\w\s]*$/.test(chunk)
      )
      .join(' ')
      .replace(/\s+/g, ' ')
      .trim()

    const content = readableText
      ? sanitizeTextForUTF8(readableText)
      : 'Unable to extract text from PowerPoint file. Please ensure the file contains readable text content.'

    return {
      content,
      metadata: {
        extractionMethod: 'fallback',
        characterCount: content.length,
        warning: 'Basic text extraction used',
      },
    }
  }
}